Controlling Outlet Position of installed NPT threaded fitting

[vonsteimel]

Greetings,
I'm looking for your opinions on the best way to control and/or call out something.
I have someone wanting to install a NPT threaded valve into a tank via a welded boss (see sketch.jpg). Ive been able to get a drawing to work for us here at the shop which was good for a dozen or so test units (see Valve.jpg) But Now the design is ready to go out to the Subcontractors for bidding but I'm questioning how to go about controlling/calling-out both the valve height and the outlet position on the tank. If the boss is at the correct length and the valve is at the correct height then the NPT thread depth will not be to spec, therefore NPT thread go/no-go gauges won't work.

I can get some leeway just by how much PTFE tape is applied to the thread in order to get it to be oriented in the right direction but not enough to leave placement to chance. The valves have to be removed before welding because the float/plunger is plastic and will melt.

I was thinking I'll probably send the contractor a valve to use a gauge or other manufacturer a gauge ourselves and send it out.
Any suggestions?
Thanks,

VS

 

[vonsteimel]

second attachment
SKETCH.jpg

 

[berkshire]

Is there any way you can use a fitting that is not so critical as to the angle of rotation such as a parallel thread with an O ring seal?
B.E.

 

[tygerdawg]

Some variant of a banjo fitting may be an alternative.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[Oldhydroman]

Unless you have a "clocked" thread on every valve then you could line up the boss for one valve and find that, at some time in the future, a replacement valve screwed into the original boss points in the wrong direction.

With tapered threads you can't dictate the tightening torque as well as the final part orientation. If you are going to allow differing thicknesses of PTFE tape so that you can achieve the final orientation with the fitting being tightened up with the right sort of torque, then you might as well not bother trying to define the boss thread orientation at all.

It doesn't look like much pressure is involved so you could try specifying a gap-filling, anaerobic, semi-hardening pipe sealant instead of PTFE tape. You could then turn the valve to the right orientation without worrying about exceeding the installation torque – leave it slack(ish) and let it stiffen up as the compound sets.

Or consider putting a shallow NPSM lock nut on that portion of the thread which remains above the boss when you have screwed the valve in most of the way and you have the orientation just right. The lock nut can then be run down the thread to tighten against the boss and hold the whole thing fast.

I would question the technique of fitting a hose clip a certain distance from the end of the tap - not all taps are the same so defining a distance from the end of the tap will not result in a consistent maximum diameter.

Also be aware that the boss can shrink when it has been welded to the tank. With parallel threaded bosses it is normal to install a solid plug before welding otherwise the fitting is very tight to install afterwards. With a taper threaded boss this technique is often ignored because the tapered fitting will still work afterwards - but the final orientation will be different after welding.

You could try changing the boss to an SAE J1926-1 port form. Then, if you have room, use a fitting such as a Parker C50MX male stud elbow. If height is a problem you could consider sinking the boss through the tank lid and welding from the underside.

The barbed hose tail can have a 37 degree flared end and connect directly to the elbow via a tube nut (actually this can be standard hose insert). Your float assembly can be brazed into the base of the fitting.

When fully assembled the hose tail can be oriented in any direction. There will be a small variation in height as you adjust the orientation: typically 0.05" per full rotation because the SAE port uses a 20 tpi thread at these sorts of sizes. This fits in with your specified height tolerance.

 

[fegenbush]

When allowable, I have used socket welded unions to control small differences in pipe length. The pipes coming in to the union can be threaded on one end and trimmed to suit and welded in the field.

 

[HowDidYouBreakThat]

In my experience using the thickness of PTFE tape applied to control the angle of an output has been dismal at best. This approach is prone to leaks, either at installation or down the line. I don't claim to be an expert, but my +25 years experience working with pipe thread tells me that this is a bad idea. If the output angle is critical, I would strongly suggest a different method of attachment.

Terry
KI6FCI

 

[racookpe1978]

How much room do you have between the edge of tank and the CL of the nozzle?

That is, can your sketch include a union between the wall of the tank and the nozzle?

 

[vonsteimel]

The valves are supplied to us. If a satisfactory solution cannot be found with the existing valve, then we'll have to go back with possible alternatives. (we have looked briefly but found nothing satisfactory and manufacturing our own is expensive)

But for now, I'm assigned to find a suitable method to use the existing valve.

HowDidYouBreakThat, you right. This is why we want to have someway to control the general direction. Then if need be, we can make minor corrections (+/- 20degrees) with the amount of tape.

New PostOldhydroman, although a torque spec is not preferred for tapered threads, it can be done. The FAA is big on torque fittings and some NPT threads do have torque specs on aircraft fittings. The lock nut I think would work... I'll look into this. The hose-clamp/tap method is only for our specific tap that we used in our shop (meant for us only); I doubt anyone would be able to replicate this with any success unless we sent them out tap or otherwise the same type/mfg/model.etc -- The valve thread is not clocked therefore the valves are not interchangeable using our method (valve.jpg), each valve is specific to its boss, via the "leave valve installed in boss and forward to inventory note". As of right now, orientation of interchangeability is not a requirement, merely that it can be removed for repair or replacement if necessary. -- A gap-filling, anaerobic, semi-hardening pipe sealant could work but it has to be fuel-proof (which almost nothing is). There are situations when the tank can be tilted up (at around 30 degrees) and when full, would completely saturate the valve, thus necessitating the need for a float valve in the first place (to prevent the fuel from pissing out the vent). Otherwise a simple roll-over valve would have been suffice.

I'll look into the SAE J1926-1 port form.

The socket welded unions still requires custom trimming; I can't seem to visualize how it could be used to install the valve given the circumstances.

I think the only requirement for distance between the edge of tank and the CL of the nozzle is that there is no over-hang. But I cannot visualize how a union between the wall of the tank and the nozzle would be a solution.

Thanks,

VS

 

[Screwman1]

There is no way to control that tap as far as the starting point of the thread, you could be off by any angle, which would require up to 359 degrees of rotation to line up. This design is not a good idea, when plumbers use tapered fittings and they want to get things to line up they just brute force the fittings into alignment and you don't want that on an aircraft part. I would go with a different sealing method, even using the setting type compound is not a high reliability solution on a fuel system fitting.

 

[vonsteimel]

Note: this is not an aircraft application.

Also, this is a purpose built vent for fuel systems purchased from the consumer market. It is sold on many racecar/dragster websites as are many other NPT fittings. Somehow they have to have a way to control the orientation with this type of thread? Even on aircraft fittings, there are many occasions when one needs to control the outlet position of an NPT fitting (even when there is a Torque spec)...

Why do the simplest things have to be so darn difficult....?

 

[SnTMan]

vonsteimel, I don't suppose its possible to install the valve thru a compression fitting?

Regards,

Mike

 

[MintJulep]

Tapered pipe threads are evil.

How about an adapter with female pipe on the top and a female flare fitting on the bottom - a nominal size or two bigger than the pipe thread. Then put male flare on the tank. Tube sticks down through the center.

 

[Oldhydroman]

Maybe the solution is to just carry on as you are, but keep a 1/4" NPT tap handy when you come to install the valve into the tank. Put on a specific length of thread tape and tighten up the fitting. If you don't like where it stops when you've tightened it up to the correct torque (plus or minus some tolerance) then note how much further you would like it to turn. Take the valve out and run the tap down the hole again.

When the tap starts to cut just turn the tap wrench through an angle similar to the correction you wanted (maybe a little less) and try the assembly again - using just as much thread tape as you did the first time.

I agree with Mintjulip: tapered threads are evil.

 

[BillPSU]

Use an SAE oring boss elbow with a barb. You will need to weld in an insert with the O-ring boss profile macined into the insert.